Deposition of xylan isolated from Pennisetum purpureum on fibres of Eucalyptus globulus and characterisation of the composition of the surface xylans by immunolabelling and enzymatic peeling

Extraction of P. purpureum xylan and its redeposition on eucalyptus pulp:

Xylans were extracted from an unbleached paper grade Soda-AQ pulp from P. purpureum (kappa number, KN 20, 18.5% xylan) using CCE at 10% consistency, 60 min and room temperature. White liquor (active alkali of 150 g l⁻¹ and sulfidity of 23%) with two alkali concentrations were used, namely: 400 and 700 kg NaOH (Merck S/A, São Paulo, SP, Brazil) per odt pulp, which corresponds to 44.4 and 77.8 g l⁻¹ NaOH concentration, respectively. At the end of reaction, the CCE liquor was removed from the slurry by centrifugation. The xylan concentrations in the two CCE filtrates were 11.1 and 14.4 g l⁻¹. These filtrates were neutralised to pH 10.5 with sulfuric acid (Merck S/A, São Paulo, SP, Brazil) and stored under refrigeration for further evaluation.

The two CCE liquors were the source of xylan to be redeposited onto an unbleached Eucalyptus globulus (Labill.) soda-AQ pulp (KN 15, xylan content of 16.5%). The xylan reabsorption was carried out in the oxygen delignification stage, which was run in a Mark V mixer/reactor (Quantum Technologies Inc., Akron, OH, USA). The E. globulus brown pulp sample was deposited at 35% consistency into the hastelloy reactor bowl and sufficient CCE filtrate plus NaOH solution (250 g l⁻¹) were added until the consistency reached 12%. The same volumes of the two CCE filtrates (11.1 and 14.4 g l⁻¹ xylans) were added to the reactor. The xylan charges to the O-stage were 5.5 and 7.2% (b.o. pulp), respectively, for the two CCE xylan extracts. The alkali charge in the O-stage was maintained constant at 2% (b.o. pulp) in both case scenarios and was provided by the 250 g l⁻¹ NaOH solution plus the alkali contained in the pH 10.5 xylan extracts. The O-stage was run at 12% consistency, 100°C, 60 min, 600 kPa, with 1.6% O₂ charge b.o. pulp weight. At the end of the reaction, the pulp was washed extensively with excess water for 24 h in running water. The O-stage gravimetric yield for the E. globulus pulp was calculated as 101.3 and 102.4%, respectively, for the two xylan concentrations. The xylan contents of the two pulps after oxygen delignification were 19.7 and 20.7%, respectively.

Extraction and characterisation of P. purpureum and eucalyptus xylans:

For analytical purposes, xylans were extracted from the P. purpureum and eucalypt brown pulps (described above) using CCE at 10% consistency, 15 min at room temperature with an alkali charge 550 kg of NaOH odt. The xylans were precipitated from the solutions by ethanol (Sigma Aldrich Brasil, Jurubatuba, SP, Brazil) addition and air-dried (Longue Júnior et al. 2013).

The compositions of the CCE-extracted and precipitated xylans were analysed. The samples were first washed with ion-exchanged water to remove the main part of the residual inorganics, followed by dialysis. Cellulose ester membranes Spectra/POR^® Biotech with cut-off 100–500 Da (Spectrum Laboratories Inc., Rancho Dominguez, CA, USA) were used. The carbohydrate compositions were determined using acid methanolysis as described by Willför et al. (2009). The fibres were also submitted to enzymatic hydrolysis as described below. The end point (48 h) was analysed for total composition, including oligomeric structures to detect the hexenuronic acid (HexUA) containing structures. The concentrations of monosaccharides – arabinose (Ara), rhamnose (Rha), galactose (Gal), glucose (Glc), xylose (Xyl) and mannose (Man) – and oligosaccharides were analysed using high-performance anion exchange chromatography (HPAEC) (Dionex ICS-3000) combined with pulse amperometric detection (PAD) (Dionex Corporation, Sunnyvale, CA, USA). Pre- and separation columns (Dionex CarboPac PA-1) were employed as described in the literature (Tenkanen et al. 1997; Tenkanen and Siika-Aho 2000; Willför et al. 2009). The standard deviations (SD) for the individual sugar amounts were Ara 1.2–3.3%, Gal 0.5–1.3%, Glc 0–1.2%, Xyl 0.3–3.9% and Man 1–2.1%. These SDs are in line with earlier published results by Willför et al. (2009) and Dahlman et al. (2000).

For MMD, the xylan samples were dissolved in alkali (4 mg sample into 4 ml of 1 M NaOH) overnight under magnetic stirring at room temperature. The preparations were filtered with 0.45 μm syringe filter and analysed using a high performance size exclusion chromatography (HPSEC) system (Waters, Milford, MA, USA) consisting of an isocratic pump (Waters 510), Waters 717 plus Autosampler with a 100 μl loop, PSS MCX 1000 and 100000 columns (100–70000 and 1000–1000000 Da), differential refractive index (RI) detector (Waters 410) and Waters 2998 Photodiode Array Detector. The columns were kept at 25°C and the flow rate of the mobile phase (0.1 M NaOH) was fixed at 0.5 ml min⁻¹. Calibration was done with Pullulan and polystyrene sulfonate (PSS) standards (Hakala et al. 2013).

Fibre surface analysis by immunolabelling:

Immunolabelling microscopy served for surface mapping of precipitated P. purpureum xylan on the eucalypt pulp fibres according to Lappalainen et al. (2004). A primary antibody was used (produced in-house), which recognises an epitope consisting of a short xylooligosaccharidic stretch carrying a 4-O-methyl-glucuronic acid residue (MeGlcA-Xyl2-3), but also polymeric xylan without cross-reactivity with other surface polymers like glucomannan. Therefore, this primary antibody is assumed to be active towards both the grass and E. globulus xylans.

Fibre surface analysis by enzymatic hydrolysis:

Two hundred milligram samples were hydrolysed using three enzyme systems: (1) combination of commercial cellulase (Econase, AB Enzymes, Rajamäki, Finland), xylanase (Ecopulp X-200, AB Enzymes, Rajamäki, Finland), mannanase (Gamanase, Novozymes, Bagsvaerd, Denmark) and β-glucosidase (Novozyme 188, Novozymes, Bagsvaerd, Denmark). The mixture was charged according to endoglucanase dosage of 50 FPU g⁻¹ at 1.6% consistency for a maximum 48 h at 40°C and pH 5 (Tenkanen et al. 1999). (2) Endoglucanase I (Cel7B) was purified from Trichoderma reesei culture filtrate according to Suurnäkki et al. (2000). Enzyme dosage of 0.25 mg protein g⁻¹ of dry pulp, 1.6% consistency, 45°C and pH 5. This enzyme has a high activity towards both xylan and cellulose. (3) Xylanase treatments were performed using xylanase purified from T. reesei culture filtrate according to (Tenkanen et al. 1992) but omitting the last gel filtration step. Xylanase dosage of 500 nkat g⁻¹ was applied at 1.6% consistency, 45°C and pH 5. Samples were taken during the hydrolyses at several time points and boiled for 10 min to inactivate the enzymes. Samples from each time point were also hydrolysed further by means of the total enzyme hydrolysis procedure to deliver the total composition as monosaccharides and analysed using HPAEC-PAD according to Willför et al. (2009) as described above for the extracted xylans.

Characterisation of the P. purpureum and eucalyptus xylans

The compositions of the isolated xylans were analysed, while the original xylan samples were purified prior to the analysis as described in the experimental, because the residual alkali in the samples interfered with the composition analysis. Methanolysis is the preferred method for uronic acids for non-cellulosic samples. One aim was to detect the compositional differences between the two xylans depending on their origin. As seen in Table 1, the P. purpureum xylan is highly substituted by Ara side groups (ara:xyl=1:15), while only traces of Ara were detected in eucalyptus xylan (ara:xyl=1:800). Thus, the presence of Ara in the eucalyptus pulps is an indicator for deposited P. purpureum xylan.

The HexUA-contents of the xylan samples were found to be 0.45 and 4.75 mg per 100 mg (corresponding to 25 and 270 mmol kg^-1) for the P. purpureum and eucalyptus samples, respectively. Thus, deposition of P. purpureum-based xylan should not significantly increase the HexUA content of the pulp.

As inorganic contaminants were removed by dialysis, the impurity of xylans is of lignin origin. The lignin content in the P. purpureum sample was lower than in the eucalypt sample. The presence of lignin was verified with MMD using ultraviolet (UV) (280 nm, lignin detection) and RI (carbohydrate detection) (Hakala et al. 2013). The average Mw of the P. purpureum and eucalypt xylans was 20700 and 16600 Da, respectively. In both cases, the lignin moiety was similar, which was probably bound to the xylan (Figure 1). The average Mw of the lignin contaminant in the P. purpureum and eucalypt xylans were 23400 and 11200 Da, respectively. However, linkages between xylan and lignin are expected to increase artificially this average data. The MMD of unpurified xylan samples was also recorded. After the purification, the solubility of the samples in alkali was insufficient and thus the analysis was not reliable because of aggregation effects. This confirms the earlier finding (Buchert et al. 1996) that alkali extracted xylans are in reality LCCs, and reprecipitation of xylan also leads to lignin content increment on fibre surfaces.

Figure 1:

Molar mass distribution of the Pennisetum purpureum (Schumach.) and eucalypt xylans detected using RI detector for the carbohydrates and UV 280 nm for the lignin impurity.

Deposition of P. purpureum xylan on eucalyptus pulp

The P. purpureum xylan was redeposited onto eucalyptus brown pulp (KN 20) during the oxygen delignification stage. The carbohydrate compositions of the pulps were determined using total enzymatic hydrolysis (48 h). The enrichment of the xylan-deposited pulps is clearly seen in Table 2. The ara:xyl ratio was also lowered by increasing xylan deposition due to the observed higher degree of Ara substitution in the P. purpureum xylan. However, the ratio was significantly lower in the eucalypt pulp (1:20) compared to the extracted eucalypt xylan (1:800), probably due to cleavage of Ara side chains during the extractions. This means that the extracted xylan cannot be regarded as a good model for eucalyptus pulp xylan. Consequently, only the distribution of Xyl rather than Ara is a reliable indicator of xylan deposition.

Fibre surface composition by immunolabelling

Immunolabelling microscopy served for surface mapping of xylan deposition. The large antibodies do not penetrate into the fibre, thus they are surface specific. The antibody applied here is able to recognise the xylan both from eucalyptus and P. purpureum. Images of the labelled samples are presented in Figure 2. In general, the antibody labelling of xylan was successful. However, as already pointed out, the method does not discriminate between the native and reprecipitated xylan. On the untreated reference, moderate labelling is visible, which appears uneven and faint with some brighter patches on fibre surfaces. Extensive labelling was observed especially on damaged fibres and fibrillated fines. This is probably due to increased accessibility of xylan at fibre kinks and dislocations. The high surface area of fines also enhances recognition by the rather large immunoglobulin molecules. In addition, fines contain higher xylan concentrations than fibres. No significant difference was seen between the two xylan deposited samples (only one of them is presented in Figure 2). Accordingly, the xylan coverage of the fibres is not even, and the labelling is partly due to the xylan-enriched outer fibre layers and partly from the unevenly distributed fines-like patches. These images are consistent with those of the previous immunochemical studies concerning the agglomeration of added xylan on wood chips and cellulose fibres (Salmen et al. 2013; Hutterer et al. 2017).

Figure 2:

Immunolabelling of xylan on eucalypt fibres.

a–c: reference sample, d–f: CCE400 with precipitated xylan. (a) and (d): bright-field images, (b) and (e): images of fluorescence emitted by the Alexa Fluor 633 labelled immunoconjugate, (c) and (f): syntheses of the two images. The arrows indicate typical sites of xylan enrichment in the two cases.

Fibre surface composition by enzymatic hydrolysis

Enzymatic peeling methods were applied according to Sjöberg et al. (2005) for the total enzymatic hydrolysis. In this method, eight time points (1, 5, 15 and 30 min, and 1, 2, 24 and 48 h) were selected for monitoring the hydrolysis process. In this time period, practically quantitative hydrolysis was reached. The idea behind this method is that fibre surfaces hydrolyse prior to the inner parts. The surface composition is represented by the first hydrolysis fractions, while the final hydrolysis results are typical for the overall composition. As expected, the proportion of xylan was highest in all samples in the beginning of the hydrolysis (Figure 3a), and also highest in the xylan-enriched samples. The slight enrichment detected in the reference samples was probably due to native enrichment of xylan in the surface layers. The point, at which 10% of the material is hydrolysed, can be interpreted as the approximated surface xylan content. Figure 3b shows the xylan contents at 10 and 90% hydrolysis levels, which are for the surface and bulk compositions, respectively. The results indicate that the xylan enrichment in the CCE 400 and CCE 700 pulps is significantly higher on the fibre surfaces compared to the average composition. Therefore, the added xylan is expected to improve significantly the paper’s technical properties.

Figure 3:

Treatment by the total enzymatic hydrolysis mixture.

(a) Xylan content (as xylose proportion in the hydrolysate) of the fibre surface (10% hydrolysis level) and bulk composition (90% hydrolysis level) by enzymatic peeling in the CCE 400, CCE 700 and reference pulps. (b) Xylose proportion of fibre surface and bulk material in the 10% and 90% hydrolysis levels from the same experiment.

The samples were also hydrolysed by pure T. reesei endoglucanase I (Cel7B) (Figure 4). This enzyme has a high activity on both xylan and cellulose (Suurnäkki et al. 2000). The hydrolysates consisted mainly of Xyl, but the degree of hydrolysis remained low perhaps due to restricted enzyme penetration because of the lack of assisting enzymes (cellobiohydrolases and enzymes with side chain cleaving activities) that are needed for the complete hydrolysis of the cell wall polysaccharides. Thus, the result probably represents the outermost surface composition. The actual level of xylan enrichment is probably lower than observed, as xylan is the preferable substrate for the endoglucanase I. The 2% and 14% hydrolysis levels were selected as the early and final stages of the hydrolysis. The pulps could be clearly differentiated according to their treatment history. However, the absolute values of the xylan hydrolysis levels are arbitrary due to the restricted hydrolysis capacity of the enzyme.

Figure 4:

Treatment by endoglucanase I (Cel7B).

(a) Xylan content (as xylose proportion in the hydrolysate) of the fibre surface by enzymatic peeling in the CCE 400, CCE 700 and reference pulps as a function of hydrolysis level. (b) Xylose proportion of fibre surface and bulk material in the 2% and 14% hydrolysis levels from the same experiment.

Hydrolysis by pure specific xylanase liberated in practice only xylan. This is the reason why the intermediate hydrolysis points did not reveal any additional information compared to the final hydrolysis time points of 24 h and 48 h, without yield increments between these points. The final levels of xylan hydrolysis were 3.0% (reference), 5.8% (CCE 400) and 6.8% (CCE 700). In line with the sample treatment, highest xylan levels were found for the xylan deposited samples and lowest for the reference. However, the xylan enrichment on the fibre surface could not be clearly visualised by means of this method.